Ecology and Epidemiology

Pollination Effects on Pearl Millet Ergot

R. P. Thakur and R. J. Williams

Plant pathologist and principal cereal pathologist, Cereal Improvement Program, International Crops Research Institute for theSemi-Arid Tropics (ICRISAT), 1-10-45/46 Chikoti Gardens, Begumpet, Hyderabad 500 016, Andhra Pradesh, India.

Journal Series Article 84 of the International Crops Research Institute for the Semi-Arid Tropics.The authors are grateful to P. R. Rao for technical assistance and to the staff of ICRISAT Information Services for the

photographs and drawings.Accepted for publication 24 June 1979.

ABSTRACT

THAKUR, R. P., and R. J. WILLIAMS. 1980. Pollination effects on pearl millet ergot. Phytopathology 70:80-84.

Pollination of inflorescences of three pearl millet male-sterile lines before absence of pollen, along with observations on the effects of pollination onor at the same time as inoculation with Clavicepsfusiformis conidia reduced stigma morphology and longevity, lead us to conclude that ergot infectionergot to less than 3% infected florets. Pollination 16 hr after inoculation in pearl millet occurs mainly through the stigmas, and that pollinationreduced ergot to 7-21% compared with 60-86% for inoculated, protects against infection probably because it induces rapid withering ofnonpollinated checks. Pearl millet pollen germinated in 1 hr when stigmas. The results help explain the increased ergot susceptibility of pearlincubated in a 10% sucrose solution, whereas C.fusiformis conidia did not millet F1 hybrids compared with traditional cultivars, and have importantgerminate in less than 16 hr of incubation. Results from inoculation of implications for ergot-resistance screening methods and for theinflorescences of male-sterile lines at different stages of flowering in the development of cultural measures to control this disease.

Additional key words: Pennisetum americanum, honeydew, protogyny, pollen-based escape mechanism.

Pearl millet (Pennisetum americanum [L.] Leeke) provides the commencement and duration of flowering events varies withstaple food for millions of people in the semiarid tropics. The cultivar and environment.average yields of traditional cultivars are estimated to be 468 kg/ha The effect of pollination. The experiments were conducted in thein India and 650 kg/ha in Africa (8). In the last 10 years fields of ICRISAT Center, during the monsoon (rainy) seasons ofconsiderable increases in pearl millet yield potentials have been 1976 and 1977.demonstrated in India through the development of FI hybrids. Inoculation and pollination treatments were made in three male-Unfortunately these hybrids have been highly susceptible to one or sterile (ms) pearl millet lines and pollen was provided by themore fungal diseases, including downy mildew (Sclerospora corresponding male-fertile lines. Individual tillers were baggedgraminicola) and ergot (Claviceps fusiformis Loveless) (2,18,21). with white parchment-paper bags at the boot-leaf stage so that theErgot is important not only because it reduces yield but also inflorescences emerged inside the bags and thus were isolated frombecause alkaloids in the fungal sclerotia are toxic to humans and pollen and inoculum.animals that eat millet products prepared from sclerotium- Inflorescences were inoculated at the protogyny stage by dippingcontaminated grain (3,5,9,11,15). Measures suggested for control them into a C. fusiformis conidial suspension of approximatelyof ergot in pearl millet include removal of sclerotia from seed prior 1 X 104 macroconidia per milliliter, prepared as a water dilution ofto planting by flotation in 10% salt solution (14), postharvest deep fresh honeydew from previously inoculated inflorescences. Theplowing to bury sclerotia (21), adjustment of date of planting to bags were removed immediately prior to inoculation and replacedavoid conditions conducive to infection (19), adjustment of fertility immediately following inoculation.levels in the crop (10), and the use of fungicides directly on the The standard pollination method used by ICRISAT pearl milletinflorescences (1,5,17,20,21). The effectiveness of these control breeders was used in this study. Bags covering inflorescences of msmethods is questionable; some are not based on experimentation lines were removed and replaced with pollen-containing bags fromor on a clear understanding of the epidemiology of the disease, and bagged inflorescences of the corresponding male-fertile lines. Theothers are not likely to be economically or technically feasible for bagged heads were pollinated by holding them horizontally andpeasant farmers. tapping the bags to disperse the pollen over the stigmas.

During the past 3 yr we have attempted to identify sources of Each of the three ms lines was subjected to seven pollinationresistance to ergot in pearl millet and have examined several aspects and/or inoculation treatments (Table 1). The single operationof the biology and epidemiology of the disease. This paper deals treatments were carried out at the maximum fresh stigma stage. Inwith investigations of the effect of pollination on pearl millet the other treatments, the first operation (pollination orergot, and we discuss the implications of the results and the possible inoculation) was carried out at the maximum fresh-stigma stage. Incontrol measures indicated by them. Part of this work has been the treatment "pollination immediately after inoculation", thepublished as an abstract (22). inflorescences were first dip-inoculated and then immediately

rebagged with bags containing pollen. Twenty inflorescences fromMATERIALS AND METHODS each ms line were used for each treatment.

The degree of ergot infection was estimated at crop maturity byFlowering in pearl millet. Pearl millet flowering is protogynous. comparing individual inflorescences with a set of standard

The long feathery stigmas emerge between the glumes 2-5 days drawings of various degrees of ergot sclerotial developmentbefore the onset of anthesis (Fig. 1,2). Stigma emergence begins (Fig. 3).near the tip and progresses toward the tip and the base. The time of The effect of flowering stage at inoculation. Tillers of two ms

lines were bagged at the boot-leaf stage and the inflorescences wereinoculated at several flowering stages from preprotogyny to

0031-979X/80/02008005/$03.0o/0 postanthesis (Table 2). Ten inflorescences were inoculated at each©1980 The American Phytopathological Society flowering stage of each ms line. Pollination was not performed.

80 PHYTOPATHOLOGY

The effect of preinoculation bagging. Several rows of pearl millet stigmas had withered (Table 2).hybrid ICH 105 were grown among rows of many pearl millet The effect of preinoculation bagging. The mean ergot incidencegenotypes during the rainy season in 1978. At intervals throughout in preinoculation-bagged inflorescences was 76%, which issix rows of ICH 105, tillers of randomly selected plants were bagged significantly greater than the mean incidence (34%) inwith parchment-paper bags at the boot-leaf stage. When the inflorescences not bagged prior to inoculation. Means per rowinflorescences were at the protogyny stage of development, 20 varied from 67-81% for preinoculation-bagged inflorescences andbagged inflorescences and 20 nonbagged inflorescences in each of from 17-46% from inflorescences not bagged before inoculation.six rows were dip-inoculated. All inflorescences were bagged The lower infection in the inflorescences not bagged prior toimmediately after inoculation. The inoculations were conducted inoculation was probably due to the protective effects ofover a 5-day period. pollination from surrounding cultivars.

Relative germination rates of pollen and conidia. Pollen and Relative germination rates of pollen and conidia. Pollenfresh honeydew conidia were collected from five pearl millet germination 1 hr after incubation varied from 7-20% and showedbreeding lines in the field between 0630 and 0700 hr. At 10-min no increase in subsequent observations (Table 3). Conidialintervals the pollen grains and the conidia from one cultivar after germination was not observed before 16 hr of incubation, and itanother were suspended in 10% sucrose solution in cavity slides steadily increased up to the final observation at 32 hr (Table 3).(10% sucrose promotes maximum pollen germination and has no Germination of pollen and conidia separately in 10% sucroseeffect of conidial germination [R. P. Thakur, unpublished]). The solution may not be the same as when they are together on theslides containing the pollen and the conidial suspensions were stigmatic surfaces, but the much more rapid germination of pollenmaintained in moist chambers at 25 C. At fixed intervals after compared with conidia and the effect of pollination on stigmaplacement in the sucrose solution, the pollen and honeydew conidia withering is consistent with the observed inhibition of ergotwere observed microscopically for germination, development by pollination.

The effects of pollination and inoculation on stigma The effects of pollination and inoculation on stigmamorphology. Plants in seven pearl millet breeding lines werebagged at the boot-leaf stage, and the inflorescences at protogynywere either pollinated, inoculated, or sprayed with distilled water. TABLE 1. Percent ergot-infected florets in inflorescences of threeInflorescences were immediately rebagged following the male-sterile pearl millet lines subjected to various pollination (poll)treatments. At intervals following the treatments, the stigmas were and/or inoculation (inoc) treatments in two experimentsexamined for morphological changes.

Infected floretsb (%)

RESULTS Treatmenta Experiment I Experiment 2

The effect of pollination. Inoculated nonpollinated in- 5141-A III-A Ill-A 5054-A

florescences developed 60-86% ergot (Table 1). Pollination prior to No inoc, no poll < 1 < I 0 0or at the same time as inoculation reduced ergot to a low level Inoc, no poll 86 65 60 60(overall means of < 3% florets infected). Inflorescences pollinated Poll, no inoc 0 < 1 0 0(overall m ns lof < mr Poll 24 hr before inoc 2 3 2 316 hr after inoculation developed more infection than those Poll 8 hr before inoc < 1 < 1 3 8pollinated earlier, but the infection was significantly less than that Poll immediately after inoc 3 4 3 1in inoculated, nonpollinated inflorescences. Poll 16 hr after inoc 8 21 7 1i

The effect of flowering stage at inoculation. Inoculation ofinflorescences when they had maximal numbers of fresh stigmas LSD (P 0.05) 4 9 6 7resulted in maximum (up to 80%) ergot incidence. Infection was a Single operation treatments and the first of double operation treatmentssignificantly lower in inflorescences inoculated at earlier flowering were done at the maximum-fresh-stigma flowering stage.stages and was less than 2% in inflorescences inoculated after all bMean of 20 inflorescences per datum.

ANTHESIS COMPLETEBEGINS ANTHESIS

POLLEN IS SHED

I5 -7 DAYS 2-4 DAYS

INFLORESCENCE 5 - 7 SEEDBEGINS TO EMERGE -"

FROM THE 'BOOT' 1-2 2 - 4 2 - 4 DEVELOPMENTDAYS DAYS DAYS

STIGMAS TOWARD COMPLETE COMPLETETIP BEGIN TO STIGMA STIGMA WITHERINGEMERGE EMERGENCE

Fig. I. Time-course of flowering events in pearl millet. Cultivar and environment can affect duration of events within the range indicated.

Vol. 70, No. 2,1980 81

morphology. Stigmas on inoculated inflorescences and checkinflorescences sprayed with water remained long, feathery, turgid,and generally white, with slight tip browing on the "sunny" side ofthe inflorescence, until the final observation 24 hr after treatment.The stigmas on pollinated inflorescences contracted considerablyand curled within 3 hr after pollination (Fig. 4); 8 hr afterpollination they were completely brown and withered. Inoculatedand water-treated inflorescences cut from field-grown plants andkept at 25 C for 54 hr in an incubator maintained completely freshlong feathery turgid white stigmas, whereas pollinatedinflorescences kept in the incubator for the same time hadcompletely brown and withered stigmas (Fig. 5).

DISCUSSION

The major factor affecting successful infection of pearl milletflorets by C. fusiformis conidia appears to be the availability offresh receptive stigmas. In the ms lines, ergot infection was greatestin inflorescences inoculated when maximum fresh stigmasoccurred and was reduced to a trace in inflorescences inoculatedwhen the stigmas had withered even in the absence of pollination.Reddy et al (17) obtained similar results with pearl millet hybridHB-1, but they concluded that the effect was due to aging of theovaries. Our results on the effects of pollination on infection, andthe effects of pollination on stigma morphology and longevityindicate that availability of fresh stigmas, and not age of the ovules,is the maj or factor that determines the success of inoculation. Fromthis, we conclude that the stigmas provide the major infection routefor the pathogen. The low levels of ergot on inflorescencesinoculated after stigmas had withered suggest that there is anotherinfection route but that it is comparatively unimportant.

There are conflicting reports on the infection path of C.fusiformis in pearl millet and other cereals. Reddy et al (17)concluded that ergot infection in pearl millet takes place mainlythrough ovary walls and rarely through stigmas or styles.Sundaram (21) stated that C. fusiformis enters millet ovulesthrough the stigmas and styles, and that direct infection throughthe ovary walls also is common. Luttrell (13) concluded on the basisof direct histological observation and inoculation studies, thatergot infection of dallisgrass occurs through the stigmas and styles.He cited several conflicting reports on the infection path of C.purpurea in cereals and concluded that the early stages in invasionof the ovary have not been adequately documented (13). Pearlmillet flowering, which is highly protogynous with large featherystigmas pushed out between tight glumes, is quite different fromflowering in wheat and barley, and thus the ergot infection processfor pearl millet also may be quite different.

Other workers have observed that pollination and / or fertilizationreduce ergot susceptibility in other cereals, and have presentedvarious hypotheses to explain the interactions (6,7,12,16,23).Although histological studies are needed to test our hypothesis forthe mechanism of the interaction of pollination and ergotsusceptibility in pearl millet, our results help explain manyobservations on the disease, and they have important implications

TABLE 2. Percent ergot-infected florets in inflorescences of two male-sterile (ms) pearl millet lines inoculated with a Clavicepsfusiformis conidialsuspension at six flowering stages

Infected floretsa onindicated ms line (%)

Flowering stage at inoculation 11 I-A 5141-A

Inflorescence emerging from boot 36 32Stigmas begin to emerge 41 31Maximum fresh stigmas 80 78Anthers begin to emerge and stigmas

begin withering 15 6Stigmas withered and anther emergence

complete < I < IAll stigmas and anthers withered < 1 2

Fig. 2. Pearl millet inflorescences at full stigma emergence (left) and later at LSD (P <0.05) 12 14anthesis (right). a Mean of 10 inflorescences per datum.

82 PHYTOPATHOLOGY

for resistance screening and control of ergot in pearl millet, been attempted by either spray or dip-inoculating openThe greater ergot susceptibility of pearl millet F, hybrids inflorescences at the fresh-stigma stage. Because flowering in all the

compared with traditional cultivars is probably caused by delayed genotypes in the screening test is not likely to be synchronous, allpollination. Unlike the traditional cultivars, pearl millet Fi hybrids but the first flowering inflorescences could be pollinated before orhave highly synchronous flowering, both in terms of tillers on a at about the same time as they are inoculated. Thus, escapes may beplant and plants in a crop. Each inflorescence is protogynous for selected as apparently resistant. We believe that this is why little2-5 days, so a period occurs when pollen is scarce and there are progress has been made in attempts to select for ergot resistance. Inmany inflorescences at the fresh-stigma stage. In addition, the F, our screening program we now bag all plants at the boot-leaf stagehybrids generally flower earlier than traditional cultivars and thus so that inflorescences to be inoculated emerge inside the bags; ie, inmore frequently flower during rainy weather which further reduces a pollen-free environment. The bags are removed briefly duringpollen availability and provides favorable conditions for infection inoculation and immediately replaced.by the ergot pathogen. Incomplete fertility restoration can occur in At least two approaches tor ergot control in F, hybrids areF, hybrids, and is another factor that would lead to decreased indicated by these results. Variability exists in the germplasm foreffective pollen availability. To save money, farmers in India often protogyny duration. It is therefore possible that a plant with muchreplant seed saved from the F1 hybrid crop. Male sterile plants reduced protogyny, or with overlapping stigma emergence andappear as segregants in the progeny and thus pollen is further anthesis, could be produced through crossing and selection. Thisreduced, should increase the probability of rapid pollination of fresh

In screening for ergot resistance the pollen-based escape stigmas. The second possibility for control is through the use of amechanism must be avoided. In the past, resistance screening has pollen donor. This would be an early-flowering cultivar, whose

TABLE 3. Percent germination over time of pollen and Claviceps fusiformis conidia, taken from inflorescences of five pearl milletbreeding lines (LI to L5)a, and incubated in 10% sucrose in cavity slides maintained in moist chambers at 25 C

Germinationb (%)Incubation Pollen Conidia

(hr) Li L2 L3 L4 L5 Mean L I L2 L3 L4 L5 Mean

1 20 14 12 7 18 14.2 0 0 0 0 0 02 20 12 13 8 20 14.6 0 0 0 0 0 04 21 12 12 8 17 14.0 0 0 0 0 0 08 21 13 11 7 17 13.8 0 0 0 0 0 0

16 .... c ............ 1 4 4 3 6 3.624 22 13 12 7 16 14.0 3 8 5 21 6 8.632 ............ 1 10 33 30 23 19.4

aLl = 700130-S-1-1; L2 = J-1623 X WC-6-S-I; L3 = EB-237-3-1-S-2; L4 = EB-137-I-I-S-8; L5 = EB-137-I-S-I.bEach datum is based on examination of 100 pollen grains or conidia.CNo observation made.

000 00 0

0

1 5 10 20 35 50 75 90Percent Ergot I cidence

Fig. 3. Standard drawings used to aid estimation of ergot incidence in pearl millet inflorescences.

Vol. 70, No. 2,1980 83

Fig. 5. Pearl millet inflorescences 54 hr after inoculation with Clavicepsfusiformis conidia (left), pollination (center), and water spray (right). Theinflorescences were cut from field-grown plants shortly after treatmernt andwere maintained in the laboratory at 25 C enclosed in parchment pape'rfbagsuntil photographed. Note the long feathery turgid stigmas on thenonpollinated inflorescences and the withered stigmas on the pollinatedinflorescence.

7. DARLINGTON, L. C., and D. E. MATHRE. 1976. Resistahce ofmale sterile wheat to ergot as related to pollination and hostgenotype. Crop Sci. 16;728-730.

8. FAO. 1976. Production year book 30:102.9. KANNAIYAN, J., P. VIDHYASEKARAN, and T. K. KANDASWAMY.

1971. Mammaliam toxicity of ergot of bajra. Curr. Sci. (BarigalOre)40:557-558.

10. KANNAIYAN, J., P. VIDHYASEKARAN, and T. K. KANIDASWAMY.1973. Effect of fertilizers application on the incidence of ergot disease ofbajra (Pennisetum typhoides B & S). Indian Phytopathol. 26:355-357.

11. KRISHNAMACHARi, K. A. V. R., and R. V. BHAT. 1976.Poisoning by ergoty bajra (pearl millet) in man. Indian J. Med.Fig. 4. Pearl millet inflorescences 3 hr after pollination (left) and water Res. 64:1624-1628.

spray (right). Note contracted and curled stigmas on the pollinated 12. LUDWIG, R. A. 1047. The ergot disease of grains and gissesinflorescence, and its control. Can, Seed Growers Assoc. Rep. 1946-47:18-21;

13. LUTTRELL, E. S. 1977. The disease cycle and funigus-hostrelationship in dallisgrass ergot. Phytopathology 67:1461-1468.

major period of anthesis would occur at the time of maximum fresh 14. NENE, Y. L., and S. D. SINGH. 1976. Downy mildew and ergot ofstigmas of the hybrid, used in a mixture with the hybrid. The variety pearl millet. PANS (Pest Artic. News Summ.) 22:366-385.would have to be resistant to ergot, or be able to escape ergot by I5. PATEL, T. B., T. J. BOMAN, and U. C. DALLAL. 1958. Anmeans of asynchronous flowering, in order not to also act as an epidemic of ergot poisoning through ingestion of infected bajra ininoculum source. In nonreplicated isolation plots at ICRISAT southern parts of Bombay state. Indian J. Med. Sci. 12:257-261.Center, we have obtained considerable reduction in ergotincidence 16. PURANIK, S. B., and D. E. MATHRE. 1971. Biology and control

of ergot on male sterile wheat and barley. Phytopathologyin a hybrid mixed with an early-flowering cultivar (Williams and 61:1075-1080.Thakur, unpublished), but the results need confirmation in 17. REDDY, K.D.,C.V. GOVINDASWAMY, andP. VIDHYASEKARAN.replicated trials. If flowering occurs during rainy weather, pollen 1969. Studies on ergot disease of Cumbu (Pennisetum typhoides)availability may still be inadequate. Thus, true resistance is needed Madras Agric. J. 56:367-377.to assure control of this disease. 18. SAFEEULLA, K. M. 1977. Genetic vulnerability: the basis of recent

epidemics in India. Part-I. Pages 72-85 in: P. R. Day ed. TheLITERATURE CITED genetic basis of epidemics in agriculture. Ann. N.Y. Acad. Sci. Vol, 287.

19. SINGH, R., and S. N. SINGH. 1969. A note on effects of different1. ANONYMOUS. 1975. Chapter VIII: 15. Progress Report of The dates of sowing hybrid bajra-1 on grain yield and incidence ot ergot

All-India Coordinated Millet Improvement Project, 1974-75. (Claviceps microcephala [Wallr.] Tul.). Madras Agric. J. 56:140.2. ARYA, H. C., and A. KUMAR. 1976. Diseases of Bajra-a serious 20. SULAIMAN, M., G. M. LUDADE, and G. S. DAWKHAR. 1966.

problem of Rajasthan desert economy. Trans. Indian Soc. Desert Effect of some fungicides and antibiotics on sclerotial developmentTechnol. and Univ. Center Desert Sci. 1:177-182. and germination of ergot on Pennisetum typhoideum. Hindustan

3. BHAT, R. V., D. N. ROY, and P. G. TULPULE. 1976. The Antibiot. Bull. 9:94-96.nature of alkaloids of ergoty pearl millet or bajra and its comparison 21. SUNDARAM, N. V. 1975. Ergot of bajra. Pages 155-160 in:with alkaloids of ergoty rye and ergoty wheat. Toxicol. Appl. Advances in Mycology and Plant Pathology. Published by ProfessorPharmacol. 36:11-17. R. N. Tandon's birthday celebration committee, New Delhi. 3(43 pp.

4. BHIDE, N. K., and U. K. SHETH. 1957. Pharmacological study of 22. THAKUR, R. P., and R. J. WILLIAMS. 1978. The effect ofinfected bajra seed. Indian J. Med. Sci. 11:892-895. pollination on ergot susceptibility in pearl millet. Abltract.

5. BRAR, G. S., J. N. CHAND, and D. P. THAKUR. 1976. Fungicidal International Congress of Plant Pathology, 03, 1978, Miichen,control of ergot of bajra. Haryana Agric. Univ. J. Res. 6:1-5. Federal Republic of termanv, 16-23 August 1978.

6. CUNFER, B. D., D. E. MATHRE, and E. A. HOCKETT. 1975. 23. WATKINS, J. E., andL. J. LITTLEFIELD. 1976. The relationship ofFactors influencing the susceptibility of male-sterile barley to anthesis in Waldron Wheat to infection by Clavicepspurpurea(Fk.)Tul.ergot. Crop Sci. 15:194-196. Trans. Br. Mycol. Soc. 66:362-363.

84 PHYTOPATHOLOGY